Five-hundred fifteen women agreed to participate in this prospective study. Of them, 2 had spontaneous miscarriage in the second trimester, 2 underwent a therapeutic termination of pregnancy (one for trisomy 21 and one for fetal congenital heart disease detected at ultrasound) while 8 women not delivering at our center were lost to follow-up. Therefore, a total of 503 women were included in the final analysis.
The maternal baseline characteristics were compared between those giving birth to a LGA neonate was (87, 17.3%) and the remnants (416) delivering normal weight infants (Table 1).
Table 1
Maternal baseline characteristics.
|
Non LGA
(N = 416)
|
LGA
(N = 87)
|
Maternal age (mean ± SD)
|
32.4 ± 4.5
|
33.0 ± 4.8
|
Low Education level (≤ 8 years)
|
52 (12.5)
|
13 (14.9)
|
Foreign women
|
52 (12.5)
|
18 (20.7)*
|
Smoking habits
|
27 (6.4)
|
4 (4.6)
|
BMI classes
Underweight
Normal weight
Overweight
Obese
Morbidly Obese
|
19 (4.5)
266 (63.9)
67 (16.1)
56 (13.5)
8 (1.9)
|
0
36 (41.4)*
29 (33.3)
19 (21.8)
3 (3.5)
|
Nulliparity
|
263 (63.2)
|
36 (41.4)*
|
Assisted reproductive conception
|
15 (3.7)
|
2 (2.3)
|
Preexisting Diabetes Mellitus
|
3 (0.7)
|
3 (3.4)*
|
Chronic Hypertension
|
15 (3.5)
|
3 (3.4)
|
Metabolic Syndrome a
|
21 (5.0)
|
7 (8.0)
|
Data are reported as numbers with percentage in brackets. |
* p value < 0.05 |
a Metabolic syndrome is defined as the presence of at least 3 of the 5 following variables: |
− HDL < 50 mg/dl |
− TG >/= 150 mg/dl
− SBP >/= 130 mmHg
− DBP >/= 85 mmHg
− BMI >/= 30 kg/m2
|
The two groups were similar for maternal age and education level, while the rate of foreign women was higher in the LGA group. Moreover, the LGA group included less normal weight while more multiparous women. A higher rate of women with preexisting diabetes mellitus was also found in the LGA group, while metabolic syndrome was similarly represented in the two groups.
Table 2 summarizes the biochemical and biophysical markers for LGA at first trimester enrollment.
Mean arterial pressure > 90 mmHg and the mean pulsatility index of the uterine artery doppler > 90th centile, were similar between the two groups as well as plasma insulin, triglycerides, and HDL. Placental and vascular markers as PlGF, inhibin A and IL-6 mean values were comparable while the MoM of PAPP-A significantly differed between groups.
Table 2
Biochemical and biophysical markers under evaluation.
|
Non LGA
(N = 416)
|
LGA
(N = 87)
|
MAP > 90 mmHg
|
116 (27.1)
|
32 (36.8)
|
Uterine Doppler PI > 90th centile
|
44 (10.3)
|
7 (8.0)
|
Insulin (µUI/mL)
|
11.7 ± 1.47
|
15.0 ± 4.35
|
Triglycerides (mg/dL)
|
107.32 ± 4.21
|
116.00 ± 9.94
|
HDL (mg/dL)
|
64.38 ± 1.1
|
62.48 ± 2.5
|
Inhibin A (pg/mL)
|
322.13 ± 16.6
|
342.92 ± 46.8
|
Interleukin-6 (pg/mL)
|
1151.05 ± 191.6
|
986.18 ± 391.7
|
PAPP-A (MoM)
|
1.40 ± 0.75
|
1.53 ± 0.86*
|
Free Beta hCG (MoM)
|
1.12 ± 0.60
|
1.01 ± 0.58
|
PlGF (MoM)
|
1.23 ± 0.50
|
1.28 ± 0.55
|
Fetal cardiac frequency > 162 bpm
|
191 (44.6)
|
31 (35.6)
|
Mean values ± SD and numbers with percentage in brackets are reported. |
* p value < 0.05 |
MAP: mean arterial pressure; MoM: Multiple of the median |
Pregnancy outcomes are reported in Table 3. No significant differences were detected as far as GDM, pregnancy induced hypertension (PIH) or preeclampsia (PE). Interestingly, the number of women who gained more weight than recommended by the Institute of Medicine (IOM) (47.1% vs 20.9%) were increased in LGA group (Table 3).
Table 3
|
Non LGA
(N = 416)
|
LGA
(N = 87)
|
GDM
Dietary treatment
Insulin treatment
|
46 (10.7%)
11 (2.6%)
|
12 (13.7%)
7 (8.0%)
|
Pregnancy induced Hypertension
|
23 (5.3%)
|
4 (4.6%)
|
Pre-eclampsia
|
6 (1.4%)
|
1 (1.1%)
|
Weight gain above IOM recommendations
|
87 (20.9%)
|
41 (47.1%)
|
Abruptio Placentae
|
3 (0.7%)
|
1 (1.1%)
|
Fetal Growth Restriction
|
6 (0.7%)
|
0 (0.0%)
|
Table 4 showed the main perinatal outcomes. While a significantly higher percentage of women with a LGA baby underwent induction of labor, the rate of cesarean section and vaginal operative deliveries was similar between the two groups. Neonatal adverse outcomes, as NICU admission, acidosis at birth or Apgar score < 7 at 5th minute were comparable.
Table 4
|
Non LGA
(N = 428)
|
LGA
(N = 87)
|
Mode of Labour
Spontaneous
Induced
|
303 (70.8%)
101 (23.6%)
|
52 (59.7%)*
35 (40.2%)*
|
Delivery
Vaginal
Vaginal Operative
Cesarean Section
|
302 (70.6%)
28 (6.5%)
98 (22.9%)
|
61 (70.1%)
3 (3.4%)
23 (26.4%)
|
Male gender
|
214 (50.0%)
|
49 (56.3%)
|
NICU admission
|
14 (3.3%)
|
1 (1.1%)
|
Umbilical a. pH < 7.2
|
23 (5.4%)
|
6 (3.4%)
|
5th min. Apgar < 7
|
6 (1.4%)
|
1 (1.1%)
|
* p value < 0.05 |
Early prediction model of LGA risk
Based on parameters available in at first trimester, a backward stepwise logistic regression was performed to identify potential predictors of LGA among 13 relevant independent variables (age, parity, pre-pregnancy BMI, preexisting diabetes mellitus, HDL, TG, insulin, PAPP-A, PlGF, IL-6, inhibin A, fetal cardiac frequency, and metabolic syndrome). The results of both univariate and multivariable analyses were reported in Table 5. At univariate analysis LGA babies were associated with multiparity (OR = 2.41, 95%CI 1.51–3.86, p = 0.001), pre-pregnancy BMI (OR = 1.08, 95%CI 1.04–1.12, p = 0.001), pre-existing diabetes (OR = 5.04, 95%CI 1.00–25.38, p = 0.050) and PAPP-A MoM (OR = 1.30, 95%CI 1.00–1.70, p = 0.051).
The final prediction model for LGA at multivariable analysis included the following independent variables: multiparity (OR = 2.8, 95% CI = 1.6–4.9, p = 0.001), pre-pregnancy BMI (OR = 1.08, 95%CI 1.03–1.14, p = 0.002) and PAPP-A MoM (OR = 1.43, 95%CI 1.08–1.90, p = 0.013) (Table 5).
Table 5
Development of the prediction model for LGA risk
|
Univariate analysis
(n = 503)
|
Multivariable prediction model
(n = 434)
|
|
OR
|
95% CI
|
p
|
OR
|
95% CI
|
p
|
Maternal Age (+ 1 year)
|
1.03
|
0.98
|
1.08
|
0.283
|
|
|
|
|
Multiparity
|
2.41
|
1.51
|
3.86
|
0.001
|
2.80
|
1.61
|
4.87
|
0.001
|
Pre-pregnancy BMI
|
1.08
|
1.04
|
1.12
|
0.001
|
1.08
|
1.03
|
1.14
|
0.002
|
Pre-existing diabetes
|
5.04
|
1.00
|
25.38
|
0.050
|
|
|
|
|
HDL ≥ 50
|
0.63
|
0.30
|
1.29
|
0.206
|
|
|
|
|
TG ≥ 150
|
1.76
|
0.96
|
3.23
|
0.068
|
|
|
|
|
Insulin ≥ 24
|
1.74
|
0.92
|
3.29
|
0.091
|
|
|
|
|
PAPP-A MoM
|
1.30
|
1.00
|
1.70
|
0.051
|
1.43
|
1.08
|
1.90
|
0.013
|
PLGF MoM
|
1.21
|
0.75
|
1.97
|
0.432
|
|
|
|
|
IL-6
|
1.00
|
1.00
|
1.00
|
0.477
|
|
|
|
|
Inhibin A
|
1.00
|
1.00
|
1.00
|
0.338
|
|
|
|
|
FCF ≥ 162
|
0.69
|
0.41
|
1.16
|
0.157
|
|
|
|
|
Metabolic Syndrome
|
1.70
|
0.70
|
4.12
|
0.244
|
|
|
|
|
The area under the ROC curve was 70.5%, indicating a satisfactory predictive accuracy (Fig. 1).
The prediction score for LGA risk was as follows:
Score = -4.565 + 1.030 * multiparous + 0.079 * BMI + 0.358 * PAPP-A MoM.
The best predictive cut-off for this score was equal to -1.378, which corresponds to a 20.1% probability of having a LGA infant. By using such a cut-off, the risk of LGA can be predicted in our sample with sensitivity of 55.2% and specificity of 79.0%.